Get all the support you need to use our reserving platform

Please note that Arius will not install and run unless you have been granted a license. 

Administrator privileges on your computer are required in order to install Arius (see installation notes for details).

Already a user? Find the latest releases and product notes here.

Term discount rates are updated yearly by several sources and are available within the Arius Stochastic module for discounting stochastic estimates. You can download the latest rate tables below.

Download the ZIP file below to your computer. Unzip the five CSV files and place them in the same location as the prior discount rate tables (CSV files). For most users this will be:

C:\Program Files\Milliman\Arius\ConfigurationData

You may need administrative rights on the computer to place files in this directory.

Once unzipped and stored in the \ConfigurationData directory, the new rates will be available in Model Options on the Term Discount tab under “Get rates from File.”

2023 Rates (for use in 2024):

• SST YE2023 Risk Free Yield Curve (TermRates_23SST.csv)
• EIOPA YE2023 Basic RFR Curves - With volatility adjustment (TermRates_23_EIOPA w volatility adj.csv)
• EIOPA YE2023 Basic RFR Curves - No volatility adjustment (TermRates_23_EIOPA wo volatility adj.csv)
• PRA YE2023 Basic RFR Curves - With volatility adjustment (TermRates_23_PRA w volatility adj.csv)
• PRA YE2023 Basic RFR Curves - No volatility adjustment (TermRates_23_PRA wo volatility adj.csv)

Data changes in 2023 YE rates

The PRA has discontinued support for the Japanese Yen (JPY) currency. JPY appeared in 2022 rates and prior, but is not included for the 2023 update.

Download the 2023 discount rates here.

‍

Arius incorporates segment names into the names of PDF files when using Export to PDF, and in Excel worksheet names and range names when using Export to Excel or API examples. Windows standard naming conventions apply; therefore, it is important to consider these Microsoft constraints when choosing your segment names.

1. Excel worksheet names (i.e., tab names) have a maximum length of 31 characters. Arius’ Export to Excel and the API examples provided with Arius use the segment names as Excel worksheet names, which will be truncated to 31 characters.

2. The following characters and words violate naming conventions for Windows file names, Excel worksheet names and Excel defined range names. They should not be used in Arius segment names:

• < (less than)
• > (greater than)
• : (colon)
• " (double quote)
• / (forward slash)
• \ (back slash)
• | (vertical bar)
• ? (question mark)
• * (asterisk)
• Also avoid using a period.
• The following Windows reserved words should be avoided as separate words in your segment names: CON, PRN, AUX, NUL, COM1, COM2, COM3, COM4, COM5, COM6, COM7, COM8, COM9, LPT1, LPT2, LPT3, LPT4, LPT5, LPT6, LPT7, LPT8, and LPT9.

3. If you intend to use your segment name in an Excel defined range name or if you will be using the Excel API examples provided with Arius, then the following rules are recommended when choosing your segment name:

• Begin your segment name with a text character, not a number or special character.
• Only four special characters are allowed in Excel’s defined range names: underline, question mark, period, and back slash. However, all except the underline character violate file naming conventions; therefore, only the underline character is recommended as a special character for use within your segment names.
• Do not begin your segment name with something that could be interpreted as an Excel cell address. For example, R2_MySegment will return an Excel error as a defined range name because R2 is an Excel cell address.
  
  Note: API examples provided with Arius use the segment name to create Excel defined names. However, “EXAMPLE4_B_BATCHGETSETDATA VBATUTORIAL.xlsm” and “EXAMPLE11_BATCHGET DataMethodReportExhibit.xlsm” allow you to choose unique range names which do not incorporate the segment name. Also, in the remaining API examples and your own VBA code, it is possible to strip all special characters from a segment name prior to incorporating into a defined range name.

4. Note that segment names are case sensitive when using the API.

‍

Note: This only applies to versions of Arius prior to v2019a.

Before version 2019a, Arius was a 32-bit application. Those earlier versions will communicate with other 32-bit applications by default, but they will need help communicating with 64-bit applications.

In this case, you can make a small change in the registry of your computer to allow the Arius API to work with your 64-bit Excel. Any changes to the registry will require someone with administrator permission on the computer and may need to be approved by your local IT department beforehand.

To update the registry to allow the Arius API to communicate with a 64-bit version of Excel, someone with administrator permissions on the computer should follow the steps below.

1. Save the attached file Arius API w 64bit Excel.zip to your computer.
2. Navigate to the file on your computer and unzip api64.reg.
3. Double-click api64.reg to run it.

Once you have followed the steps above, you can delete api64.reg and the ZIP file from your computer.

Download the registry update here.

‍

The methods I typically use require data inputs that I don’t see listed under the data node in Arius. Can these be customized within the system? What about other exhibits, methods, and reports?

The Object Library lists all of the data arrays, exhibits, methods, and reports that are built into the file. You can also build your own data tables, exhibits, methods, and reports from within the Object Library. Refer to the Arius Deterministic Formula Functions user document for instructions, examples and a complete list of functions available for use in Arius formulas (Help | User Documentation | Formula functions).

‍

I have created some user-defined tables in my Object Library. How do I make these available for use within my collections?

Arius Collections are the groupings that organize your tables and graphs into an efficient workflow on the main interface of the software. The left-hand navigation panel allows you to customize and organize these collections. For example, if you right-click, you can see options to add/edit/delete/rename collections or add/delete/rename folders. You can also drag items to reorganize or reorder collections and folders.

The system provides dozens of collections out of the box and you can also create your own. To create a new collection, you can either copy an existing collection and make modifications or start from scratch. Refer to Tips for Working with Collections for guidance (Help | User Documentation | Collections of tables).

I am performing an interim analysis. My last diagonal is a partial period and my most recent exposure period is not fully earned. Can Arius adjust for this?

The Project Settings | Data Structure tab contains a field called "Length of Last Calendar Period." This field defaults to the length of the development period (e.g., 12 months, assuming the last diagonal is a full period); however, it is also possible to import or enter a partial diagonal of data to perform interim analyses (e.g., < 12 months). The system will recognize that the last diagonal is a partial period and will automatically (a) exclude the last diagonal of age-to-age factors from development exhibits and (b) include rows for selecting interpolated development factors on the bottom of all development exhibits.

All of Arius' methods that use development factors use the IfInterpolate function to switch between full period and partial period calculations. If the last diagonal is a partial period, the IfInterpolate function is TRUE and will, therefore, return the first parameter (e.g., Interpolated Selected Development Factors) and if FALSE, will return the second parameter (e.g., Selected Development Factors). All ultimates are projected to a full period ultimate for comparison purposes.

Lastly, in order to adjust for unearned exposure periods, the system prorates the ultimate loss in the "Summary of Reserve Estimates" reports based on the "Proportion Earned" array under Data | Assumptions (in Object Library if not in left navigation pane). The system calculates the adjustment based on your selection from several system-provided earning patterns – uniform, based on exposures, based on premiums – or you can create a user-defined array for something more unique to your data.

‍

I am trying to match interpolation results from Arius to my Excel spreadsheet where I have interpolated my factors using a linear curve fit, but my first factor does not match.

The interpolation approach used in Arius is based on the Average Date of Loss Approximation method. The key to this approach is in the handling of partial periods, i.e., the less than 12 month factor for an accident period or less than 24 month factor for a policy period. This method makes two adjustments: it first adjusts for the maturity of the partial period and it then adjusts for the unearned exposure of the partial period. For example, suppose you have a 3 month accident period evaluated at the end of 3 months. The difference between the evaluation date and average accident date would be 1.5 months. This would be equivalent to a 12 month accident period with an average accident date of 6 months, evaluated at 7.5 months. The system then multiplies this 3 month-to-ultimate factor times a factor of 4 to account for the remaining 9 months of exposure in order to project a full year ultimate. Refer to the Interpolation user guide for additional details and examples (Help | User Documentation | Interpolation).

Arius allows you to enter a scaling factor that indicates to the system that the amount fields were entered in, say, thousands (e.g., scaling factor = 1000). This scaling factor is important so that particular averages can be displayed with appropriate precision. For example, the system calculates paid severity as (Cumulate ("Paid Loss")/Cumulate ("Closed Claims")) x "Scaling Factor." The scaling factor currently only applies to the Deterministic module and can be found in the Object Library under Data|Assumptions or by pressing the "Source Data" button from any of the tables within the Claim Severity collection, where the scaling factor is used extensively.

I’d like to extrapolate my development factors to select a tail factor, but the options under the Tail Factor Analysis tool are grayed out.

While working within any development exhibit within Arius, you can invoke the Tail Factor Analysis tool. This tool is active and available on the tool bar once you have selected a row of factors to which to fit the appropriate curve(s). To use the tool, do the following:

1. Open your development exhibit.
2. Select a row of factors to which you would like to fit the curve by clicking on the row label in the left panel.
3. Right-click and select Tail Factor Analysis (or select the Tail Factor Analysis icon from the far right of the table toolbar) and select New.

Arius will fit a curve to your selected data using a least-squares regression approach. You can define how far into the future to extend the calculation, modify or omit factors, and graph the resulting curve fits. The results are then included back on the original exhibit for those curves you select to "Show on Exhibit" (see the check boxes above the graph). To zoom in on the graph, right click on the graph and select to "Enable Zoom." Refer to the Tail Factor Analysis user guide for additional details (Help | User Documentation | Tail factor analysis).

‍

The box-whisker plots available from any development exhibit within the deterministic module provide a visual representation of the variability of the age-to-age factors at each age of maturity. Statistics assumes that your observations are clustered around some central value, so the "box" represents the interquartile range (IQR) or the range of factors from the 25th to 75th percentile, and the horizontal line represents the median value. Lower and upper "fences" help identify potential outliers; the fences are calculated as a multiple of 1.5 times the length of the IQR below the 25th percentile and above the 75th percentile. The plot’s whiskers extend to the lowest observation within the lower "fence" and the highest observation within the upper "fence." Any observations outside the "fences" are marked as potential outliers. See the figure below.

Possible outliers are depicted by orange markers. To see the effect of excluding an outlier from your statistics, highlight the corresponding age-to-age factor on your exhibit, right-click, and select "Exclude Factor(s) from Exhibits." This strikes through the factor on the table, automatically excludes the factor from all statistics calculations below, and adjusts the box-whisker plot accordingly.

The green markers are your selected factors. You can identify potential bias in these selections by observing for patterns in where your selection points fall within the various graphs.

Issue: Values used in the tail factor analysis (TFA) calculations may be incorrectly excluded if both of the following are true: (a) you manually edited the TFA, and (b) values were excluded, either by the system or manually, before the data changed.

Workaround: Always review and confirm TFA values as part of your analysis process. The workaround to this issue is to manually uncheck values that are incorrectly marked for exclusion during this review, or you can simply rerun TFA.

Issue: User-defined exhibits and methods with table IDs 801, 802, 803, and 804 conflict with hidden system tables and cause the following issues: (a) these methods can’t be selected as methods within formula-driven assumption arrays and (b) the exhibit type is not saved for these exhibits. Arius v2021b no longer allows you to create tables with these IDs, though tables with these IDs could already be present in existing projects.

Workaround: If a user-defined exhibit or method exists with a table ID of 801, 802, 803 or 804, go to the Object Library, copy this table and enter a new table ID.

Issue: When creating a new segment by copying a segment from another Arius project, external factors are not included in the data that is copied in.

Workaround: When copying segments from one project to another and the incoming segment includes external factors, you will need to remap those external factors in the new segment.

Issue: The batch processing feature for Export to PDF within Arius Enterprise currently only exports Deterministic tables. If a Stochastic table or graph is included in the print list, you will get the error message Failed to Export to PDF: Object reference not set to an instance of an object on the AUTOMATION | STATUS page.

Workaround: You will need to remove these objects from your print list in order for your job to successfully complete.

‍